5 Concepts in Statistics You Should Know | Data Science Interview

Поделиться
HTML-код
  • Опубликовано: 20 ноя 2024

Комментарии • 32

  • @WebsterLincoln
    @WebsterLincoln 2 года назад +17

    I would describe that as a positively skewed normal distribution, not an exponential distribution. Also, it's the 68-95-99.7 rule

  • @abdallahelmoctar7635
    @abdallahelmoctar7635 Год назад +3

    Such a simple and straight forward refresher. I'm grateful for your work

  • @asadhasnainbaqri5455
    @asadhasnainbaqri5455 Месяц назад

    significance level is the probability of making type 1 error, which is rejecting null hypothesis when it was true

  • @annizheng5289
    @annizheng5289 4 месяца назад +1

    Nice for having a brush up! Thank you!

  • @shir0tei
    @shir0tei 2 года назад +2

    Thanks for the video! I The correlation formula is wrong though, the covariance is the numerator divided by n.

  • @RedShipsofSpainAgain
    @RedShipsofSpainAgain Год назад +8

    11:16. I think you have a typo: The Normal distribution should be 68-95-99.7%, not 65-95-99.7%

  • @mahmutozmen1261
    @mahmutozmen1261 2 года назад +2

    Thanks for such a great content and your effort. Would you mind explaining further why you think that mode = median? Since this graph seems like a positively skewed graph, I though mode is around 3, median 4 or 5 and mean between 6 and 10.

  • @Foba_Bett
    @Foba_Bett Год назад +1

    I am binge-watching your channel ! 😎
    In the correlation section - why not just straight up remove the outliers? 🤔

    • @gaboqv
      @gaboqv Год назад

      that's what he is telling with a fancy name, you will use quartiles to confirm which of the points are outliers

  • @basmaelkhamlichi8223
    @basmaelkhamlichi8223 2 года назад +8

    Hypothesis testing and P value nicely explained, thank you!

  • @cosystudy55
    @cosystudy55 Год назад

    Could you mention tools used to design and present your slides thanks!!!

  • @dreamingaparisdream3178
    @dreamingaparisdream3178 2 года назад +4

    For the normal distribution, is it 66-95-99.7 rule or 68-95-99.7?

    • @TheNIK21HIL
      @TheNIK21HIL 2 года назад +1

      it is 68% within 1 SD. it must be a typo on Dan's end. The graph though does represent it correctly.

    • @ASHISHDHIMAN1610
      @ASHISHDHIMAN1610 2 года назад

      @@TheNIK21HIL yeah typo

  • @jcokonkwo
    @jcokonkwo 2 года назад +4

    I definitely appreciate the explanation then the applied DS examples right after. Thank you!

  • @ayushmathur5984
    @ayushmathur5984 4 месяца назад +1

    in #2 at 4.30 mins how he told that mean,median and mode value can anyone explain please

  • @jacksun7999
    @jacksun7999 6 месяцев назад

    6:43 should the numerator be cov(X,Y)? Seems there is a 1/(N-1) term missing.

  • @benxneo
    @benxneo 2 года назад +2

    could you give me ideas for data science projects that deliver value to businesses

  • @AllieZhao
    @AllieZhao 2 года назад

    These are crucial concepts. Thanks

  • @pal999
    @pal999 2 года назад

    If you're using a real world example, you shouldn't "ASSUME" the SD to be something. Can you find out how it's determined in real world?

  • @dreamingaparisdream3178
    @dreamingaparisdream3178 2 года назад

    Also where is the link for Meta Statistical Interview questions video please?

  • @stanislavdidenko8436
    @stanislavdidenko8436 2 года назад

    2pm - poisson distribution

  • @anirbansarkar6306
    @anirbansarkar6306 Год назад

    Can you help me understand on what basis have you assumed population standard deviation to be 20?

  • @bandai2
    @bandai2 2 года назад

    could you also use Spearman Correlation if you have outliers in your data?

  • @HarryPotter-st2cn
    @HarryPotter-st2cn 2 года назад

    Great content. Is non-normal distributions listed separately to put emphasis on it? I believe it will be included within the concept of the overall distributions

  • @BrianSalamone
    @BrianSalamone 8 месяцев назад

    1:08 8 hours a day in Facebook????? What is the X at the bottom?

  • @michaell9804
    @michaell9804 2 года назад +4

    You failed to mention bayes theorem and binomial distribution which is used here just as heavily as normal distribution particularly when quantifying the probability distribution of the accuracy of unsupervised learning models. This video is not comprehensive at all

    • @Omegageekk
      @Omegageekk 2 года назад +13

      If you thought a video titled “5 concepts in statistics you should know” would be a comprehensive breakdown of literally every stats concept you need for data science, then I have a bridge to sell you.